It's a lot more nuanced than that as well with the US. A very small portion of the US population lives in a location with a murder rate at 3.9 or higher, because murder in the US is hyper concentrated (a few areas of Chicago have more murder annually than all of Japan). People like to pretend the US is a wasteland of murder, when in fact the very extreme majority of people in the US live in areas with murder rates comparable to Canada.
Of course urban areas are radically more dangerous when it comes to murder. The murder rates in US cities are drastically higher per capita than in rural and suburban areas, that isn't even remotely up for discussion, it's a fact that stretches back to the beginning of US data on murder.
I'm not sure why you emphasize that murder rates are per capita, that proves the point you're trying to deny. The per capita murder rates in cities are far higher. The very high city per capita murder rates in the US are the reason the US murder rate is so high.
The point being made is that you're cutting out the bad parts in the US in order to make the comparison, but not cutting out the bad parts in the comparitor. Rural murder rate in the US is less than combined rural + urban murder rate elsewhere? Quelle surprise.
Similarly, by cutting out the urban parts of the US in your comparison, you're cutting out 70% of the US population.
Is murder in other countries not concentrated? I can draw circles over a few areas in Japan and say "these areas here have 100% of murders, the rest of the country has a murder rate of zero".
Japan does not have anything even close to what would be considered a high-crime US neighborhood, and the little crime there is fairly even scattered around the country:
If you step outside of the major US cities, the murder rate per capita plunges. That variance in the US is far beyond what you see in other developed nations (eg Britain, Sweden, Finland, France, Belgium, Germany, Japan, etc). It's so bad that if you just brought the three dozen worst urban sub-areas down to normal as compared to the rest of the city in question, you'd reduce the total US murder rate more than a full point.
>if you just brought the three dozen worst urban sub-areas down to normal as compared to the rest of the city in question, you'd reduce the total US murder rate more than a full point
>three-dozen
That seems like a lot of areas to say "just". If it's that easy to bring it down, do it in actuality rather than coming up with hypothetical scenarios where the US doesn't have such a high murder rate.